Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[strontium90]

Hi again Walter
I was so sick of it that I e-mailed Ian LIvingstone...To my surprise, after an hour I got a phonecall on my mobile from his office...The (English) girl was extremely helpful, but thinks that there may be a capped speed on my line...she is contacting wholesale and will get me back to where I was (so to speak) but I will have to wait 3 days for the line to stabilise with DLM etc...
She also gave me a direct phone number to ring her back if the problem persists...
Alex

[Oranged]

but I will have to wait 3 days for the line to stabilise with DLM etc...

But this is exactly what Walter and I have been telling you.

If it was a fixed 2meg connection the DLM would not be applied, it would show a 2272kbps downstream sync and the SNRM would be significantly higher than your 6.8dB.

Looking at the last SNRM result it would appear that the noise occurrence has gone away and the IP Profile will begin to increase.

[strontium90]

Ok guys...Guess I have to be patient,,,will give it a few days with the netgear running 24/7....thanks for the replies..

[strontium90]

Sorry to bother you all again, but is this norma,l for a BT speedtest on the Dsl Max "up to 8 meg" product?

[attachment deleted by admin]

[waltergmw]

Hi again Alex,

Yes, this is to be expected for the time being, but the IP Profile should incerease before too long, unless you are still experiencing noise bursts.

In normal circumstances a sync speed (i.e. connection rate) of 4064 kbps should result in an IP Profile of 3,500 kbps.

Kind regards,
Walter

[strontium90]

I had an ip profile yesterday of 3500kps so was pretty happy....about 30mins ago there was a cloud-burst...no thunder or lightning, just heavy rain...
Th snr fluctuated a bit and suddenly there was a re-sync....does this merit a visit by Open-Reach, and more importantly, will it be free as the problem is obviously outside

Am really sick of this...Another 4-5 days waiting on the cursed DLM

[waltergmw]

Hi Alex,

I suspect you won't get very far trying to get anyone interested in a single event.
There is a chance that you have a corroded crimp joint somewhere and the rain, or strong wind, has probably disturbed your connection.
It could however also be the result of distant thunder.
Running and recording Routerstats continuously is probably the only way you can provide the evidence to encourage further investigation by your ISP and then BT Openreach.

Kind regards,
Walter

[strontium90]

Everything had been working fine...had my ip profile back to 3500kps since last Thursday....looked at my RouterStats this morning and see I've had a re-sync event at 3am...couldn't work out why as snr was 5.5dB at the time...
However from the telnet tab, I've noticed they've (DLM) changed my interleave depth on the downstream, from 32 to 1....hence more error prone..
Why would it do this?

[waltergmw]

Hi Alex,

can you post your Routerstats pictures for that time?

Kind regards,
Walter

[strontium90]

Hi Walter..
On closer inspection it seems the re-sync took place at 1am...Unfortunately, I don't know when they changed the interleave, but I'm sure the accumulation of errors caused it...I have the noise margin and the sync and will post the telnet page from now
I have resynced this morning to bring me back to my normal 4000kps sync speed
Alex

[attachment deleted by admin]

[waltergmw]

Hi Alex,

The lines look inconclusive to me in that there is no obvious cause for the re-sync.
Has anybody else her any comments / ideas ?
You could perhaps try reducing the sample time to see if another event can be trapped with more detail.

Kind regards,
Walter

[roseway]

The most likely cause of the resync was (I think) a brief burst of interference which happened between the Routerstats sampling intervals so didn't get recorded. Something electrical in the home perhaps, such as a fridge thermostat.

[sevenlayermuddle]

 

Routerstats is a very useful tool, but we need to remember that it draws its graphs by joining up the dots from a series of samples.  If a momentary noise spike breaks out between samples, it can kill the connection, and maybe even re-establish it, by the time routerstats takes it's next sample so the noise spike goes unreported.

The fact that the connection speed hadn't changed by much also suggests that, by the time it reconnected, the noise had abated again, which is further evidence that it was a short spike.

I see Eric beat me to the post, but I'll post anyway

[strontium90]

Thanks Guys..
What I don't understand is why (I'm with BT) did they switch interleave off?...see above telnet sample
Surely my errors are going uncorrected and this is what caused the re-sync

[sevenlayermuddle]

Thanks Guys..
What I don't understand is why (I'm with BT) did they switch interleave off

Is there any chance your ISP has initiated a CCSNR (Customer Controlled SNR) process, perhaps in mistaken the belief that's what you wanted them to do?

I'm not 100% clear about CCSNR, but I think that in essence it  just resets all the initial conditions line and then re-initiates the ten-day training.  That would mean your target SNRM will change back to 6dB and, I assume, interleaving would also be turned off.  We tend to think of DLM is just changing the target SNRM, but actually when DLM stabilises a line, the first weapon in its armoury is to enable interleaving,  and only if that fails to stabilise does DLM start increasing the target margin.

If CCSNR has been initiated then a 'hands off' approach is best for a few weeks to let DLM sort things out.  Don't tweak the margin, and avoid any manual resyncs, either of which could confuse DLM.


edit: added 'target'